Performance of Nile Tilapia and vegetables Grown in Different Aquaponic Volumes
Metadata
- Cite key: barbosaPerformanceNileTilapia2020
- Item type: Journal Article
- Authors: P. Barbosa, J. Povh, A. Silva, A. Ventura, G. Stringhetta, L. Laice, A. Oliveira, T. Carvalho, R. Corrêa Filho
- Affiliation: Faculdade de Medicina Veterinária e Zootecnia, Universidade Federal de Mato Grosso do Sul (UFMS), Campo Grande, MS, Brazil (all authors except Laice Menes Laice); Agriculture Division (DivAG), Institute Higher Polytechnic of Manica, Campus of Matsinho, Chimoio, Mozambique (Laice Menes Laice)
- Journal: Journal of Agricultural Studies 8(4) (2020) 497-506
- Date: 10/2020
- Date added: 2026-07-13
- DOI: 10.5296/jas.v8i4.17598
- Funding: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brazil (CAPES), Finance Code 001; Fundação Universidade Federal de Mato Grosso do Sul (UFMS/MEC), Brazil; Fundação de Apoio ao Desenvolvimento do Ensino, Ciência e Tecnologia do Estado de Mato Grosso do Sul (FUNDECT), Doctoral fellowship, Call no. 03/2017
- URL: https://doi.org/10.5296/jas.v8i4.17598
- PDF:
Barbosa et al. - 2020 - Performance of Nile Tilapia and vegetables Grown in Different Aquaponic Volumes.pdf
Opinion
A small, tightly controlled randomized-block comparison with a genuinely useful design question (does fish-tank volume alone matter, holding density and vegetable-bed volume constant?), but the paper’s own numbers undercut the ability to trust some of the fine detail — the 40-day vs 30-day duration slip in the daily-weight-gain formula, and the fact that “final weight” and “biomass gain” are reported only as tank-volume-normalized densities (kg/m3) rather than the absolute per-fish or per-tank figures a reader would want for cross-study comparison. No hydroponic control and no nitrate/tissue data limit how far this can be used outside the narrow volume-ratio question it set out to answer. Worth citing specifically for the volume-ratio finding, not as a general performance benchmark.
Abstract
The objective of this study was to evaluate the performance of Nile tilapia and the lettuce, watercress, and arugula vegetables in a gravel-bed aquaponic system using the aerated fish tank-to-vegetable growing bed volume ratios of 1:1 and 1:2. Each experimental unit consisted of an aquaponic module composed of two containers: one to allocate the vegetables (500-L tank with gravel, measuring 2 m2) and another to stock fish at the ratios of 1:1 (500-L vegetable:500-L fish tank) and 1:2 (500-L vegetable tank:1000-Lfish tank). The experiment was laid out in a randomized-block design with two treatments (500-L fish tank × 1000-L fish tank) and three blocks (periods). The performance of the Nile tilapia and of the vegetables did not differ significantly in response to the increasing fish tank volume, with the fish and the vegetables exhibiting good performance results in both volume ratios. In conclusion, in aquaponics, the fish tank-to-vegetable tank volume ratios of 1:1 and 1:2 are equally effective for the production of Nile tilapia and the lettuce, watercress and arugula vegetables.
Summary
The authors built matched gravel-bed (media-filled) aquaponic modules stocking Nile tilapia at a fixed density (4 kg/m3, 12 fish/m3) while doubling only the fish-tank volume (500 L vs 1000 L), keeping the 500-L vegetable growing bed (co-planted with lettuce, watercress and arugula) and planting density constant across both arms. Over a 40-day trial run as a randomized-block design (three periods as blocks), they measured fish growth/survival/FCR, water-quality parameters, and per-species vegetable leaf/root/biomass weights. No statistically significant differences (Student’s T test, P>0.05) were found between the 1:1 and 1:2 volume ratios for any fish or vegetable performance variable, or for water quality. The authors conclude both ratios are equally effective, though they note vegetable biomass means were numerically (non-significantly) higher at the larger 1000-L fish-tank volume and suggest this may warrant testing larger vegetable-bed volumes as well. The study has no hydroponic control arm and does not report absolute (non-density-normalized) fish weights, tissue nitrate, or water nitrate.
Experiment data
- Location: Fish Farming sector, Universidade Federal de Mato Grosso do Sul (UFMS), Campo Grande, MS, Brazil
- Design: Randomized-block design, 2 treatments (500-L vs 1000-L fish tank, at fixed 500-L vegetable bed) x 3 blocks (periods); no hydroponic-only control
- Replicates / n: 3 (blocks)
- Duration: ⚠️ 40 days named as the experimental period (Methods, p.500), but the same paragraph’s daily-weight-gain formula divides by 30 days — see Extraction notes
- Organisms: Nile tilapia (Oreochromis niloticus) / Lettuce (Lactuca sativa, crisp) / Watercress (Rorippa nasturtium-aquaticum) / Arugula (Eruca sativa)
- Statistics: ANOVA (treatment + block, fish initial weight as co-variable), GLM procedure, SAS 9.0; Student’s T test, P>0.05 = not significant
- Feed Conversion Rate (FCR): 1.68 (1:1) vs 1.50 (1:2), P=0.70
- Fish survival rate: 95.64% (1:1) vs 96.01% (1:2), P=0.58
- Total vegetable biomass: 3.44 kg/m2 (1:1) vs 4.88 kg/m2 (1:2), P=0.51
Design and volume ratios
This paper: Two aquaponic-only treatments differing solely in fish-tank volume (500 L = ratio 1:1, or 1000 L = ratio 1:2) while vegetable-bed volume (500 L, 2 m2 gravel bed) and fish density (4 kg/m3, 12 fish/m3) were held constant. Total system volume was therefore 1000 L (T1) or 1500 L (T2) including the shared 500-L vegetable tank. No hydroponic control arm exists — this is a within-aquaponics volume-ratio comparison, not an AP-vs-HYD comparison.
Compared with:
- todo Babatunde et al. 2019 — gravel-based substrate filter aquaponic co-production of amaranthus and tilapia; found tilapia performance decreased with increasing stocking density (100 to 250 fish/m3), a different manipulated variable than this paper’s fixed-density/varied-volume design (p.502).
- todo Lennard and Leonard 2006 — comparison of gravel bed, floating and NFT hydroponic sub-systems in an aquaponic test system; better biomass/yield in gravel vs NFT/raft (p.499).
Fish performance
This paper: Final weight, weight gain, daily weight gain, final biomass, biomass gain and apparent feed conversion of Nile tilapia did not differ significantly (P>0.05) between the 500-L and 1000-L fish-tank volumes (Table 2, p.501). All fish-performance values in Table 2 are reported density-normalized (kg/m3), not as absolute per-fish or per-tank figures — see Extraction notes for why this leaves several trials.csv columns NR.
Compared with:
- todo Moraes et al. 2009 — Nile tilapia in net-tanks: biomass gain 0.476 kg, FCR 1.59 [secondary comparison figures cited by this paper, p.502].
- todo Assano et al. 2011 — Nile tilapia in fish ponds: daily weight gain 2.37 g/day, FCR 1.3 [secondary, p.502].
- todo Lima et al. 2015 — Nile tilapia in biofloc technology (BFT) aquaponics: daily weight gain 2.16-2.36 g/day, FCR 1.38-1.62 [secondary, p.502].
Vegetable performance
This paper: Leaf, root and biomass weights of lettuce, watercress and arugula did not differ significantly (P>0.05) between fish-tank volumes (Table 3, p.502). Non-significant mean values were numerically higher for all three vegetables and total biomass at the larger (1000-L) fish-tank volume, which the authors speculate could reflect more available nutrients at larger volume, but call for further study with larger vegetable-bed volumes too (p.503-504).
Compared with:
- todo Sace and Fitzsimmons 2013 — NFT aquaponics with Nile tilapia (250-L tanks) and freshwater prawn polyculture; whole-lettuce weight 75.5 g over 108 days, lower than this study’s reported “80 and 121 g” (attribution to 1:1 vs 1:2 unclear, see Extraction notes) (p.503).
- todo Pinho et al. 2017 — BFT raft aquaponics lettuce varieties (red 0.56 kg/m2, butter 1.94 kg/m2, crispy 1.36 kg/m2) vs clear-water system, 500-L fish tanks, 21 days (p.503).
- todo Lennard and Ward 2019 — NFT aquaponic vs hydroponic plant growth comparison; whole-arugula weight 10.7 g (42 days) close to this study’s “14 g”; arugula found less adapted to aquaponics than lettuce, with values ~3x lower than hydroponics (31 g) (p.503).
- todo Nhan et al. 2019 — raft aquaponics watercress at 180 plants/m2, final biomass 0.460 kg/m2, close to this study’s watercress biomass (p.503).
Water quality
This paper: Temperature, pH, dissolved oxygen, total ammonia nitrogen, ammonia (NH3) and nitrite were all statistically similar (P>0.05) between the two fish-tank volumes (Table 1, p.501) and described as within the adequate range for tropical fish citing Boyd (1998). No nitrate (NO3/NO3-N) data were collected or reported anywhere in the paper.
Linked claims
- Increasing fish-tank volume alone does not improve aquaponic fish or vegetable performance when density and vegetable-bed volume are held constant
- Gravel media-filled bed aquaponic systems support co-production of multiple leafy vegetable species
Citations to chase
- todo Babatunde et al. (2019) — gravel-based substrate filter aquaponic tilapia density effects
- todo Sace and Fitzsimmons (2013) — NFT aquaponics with Nile tilapia and freshwater prawn polyculture, lettuce productivity
- todo Pinho et al. (2017) — BFT effluent aquaponics vs clear water, lettuce varieties
- todo Lennard and Ward (2019) — NFT hydroponic vs aquaponic plant growth rate comparison
- todo Nhan et al. (2019) — stocking density effects on watercress in aquaponic recirculating system
- todo Lennard and Leonard (2006) — gravel bed vs floating vs NFT hydroponic sub-systems comparison
- todo Moraes et al. (2009) — Nile tilapia net-tank performance with different commercial feeds
- todo Assano et al. (2011) — Nile tilapia pond performance with different protein sources/levels
- todo Lima et al. (2015) — Nile tilapia in biofloc technology (BFT) at different stocking densities
Extraction notes
⚠️WARN-MATERIAL — Fish trial duration. Methods (p.500) names the biometric-measurement window as “(40 days) of the experimental period,” but the same paragraph’s daily-weight-gain formula divides by “30 days”: ”[(Final weight, g − Initial weight, in g)/30 days]”. Every other reference to trial length in the paper (framing of Methods, Abstract, Conclusion) is consistent with 40 days; the 30-day figure appears only inside this one formula. Recorded 40 days as Fish trial duration (days) in trials.csv (flagged with ⚠️ prefix in the cell); the 30-day divisor is very likely a drafting error but cannot be confirmed from the text alone. Table 2’s reported Daily weight gain values (2.55 g/day at 1:1, 2.77 g/day at 1:2) are recorded as directly stated, not recomputed under either duration. Affects: interpretation of the daily-weight-gain rate’s true basis only.
⚠️WARN-CHECK — NO2-N basis. Table 1 (p.501) labels the variable “Nitrite (mg L-1)” using the ion notation “NO2-”, not explicitly “NO2-N” (nitrite-nitrogen). The paper never states whether the reported mg/L basis is the NO2- ion or its N-equivalent (differ by a factor of ~3.28). Recorded as stated (0.05 at 1:1, 0.09 at 1:2) with no conversion applied. Flagged for user verification.
⚠️WARN-CHECK — Whole-vegetable weight attribution. Discussion (p.503) states lettuce whole-vegetable weight “(80 and 121 g)” when comparing against Sace and Fitzsimmons (2013), and a whole-arugula weight of “14 g” comparing against Lennard and Ward (2019). Neither passage states which figure belongs to the 1:1 vs 1:2 treatment, nor whether 14 g is a pooled/overall mean across both. The paper’s tables consistently list the 1:1 column before the 1:2 column, which would suggest 80 g = 1:1 and 121 g = 1:2, but this is inferred from formatting convention rather than a stated label. Both trials’ “Plant fresh weight” cells in trials.csv are recorded NR rather than guessed.
[not reported] (grouped by field): Fish Category; SGR; N, P, K (feed elemental composition); % of body weight (daily ration); Fish size initial/final (g); Total Feed (kg, absolute); Fish biomass created (kg, absolute); Fish weight gain (cumulative, g); Water recycle (L/min); Water type; Water classification; Daily Water exchange rate (%); pHOptimal; FUE AP; WUE; EC; NO3-N (water nitrate — not measured at all in this paper); Plant Category; Days Plant after transplant; SPAD; Plant height; Leaf count; Plant fresh weight (see CHECK above); Plant dry matter; Tissue nitrate AP (no tissue nitrate measured at all); Lat/Long (Campo Grande, MS coordinates not stated in the paper and not filled from outside knowledge, per the prime directive); Average room Temperature; Iron supplemented; Remineralization; pH Buffers; Climate control; Artificial Lighting.
NA (not applicable — no hydroponic control arm exists in this study): FUE HYD; Tissue nitrate HYD; HYD.
[unclear]: whether the urea fertilizer addition described as being used “to stabilize the aquaponic systems” (p.500) was a one-time pre-trial cycling step or continued during the 40-day data-collection period — the paper gives no timeline, so Nutrient supplemented is recorded Y with this caveat in trials.csv remarks.
NO COLUMN items (recorded in Experimental Remarks in trials.csv, no matching schema column):
- Ammonia (NH3, un-ionized), Table 1: 0.02 mg/L both treatments, CV 18.84%, P=0.42.
- Coefficients of variation (CV%) reported for every Table 1-3 variable — not converted to SD (would be derivation); full CV list recorded in remarks.
- Fish “Final weight,” “Final biomass” and “Biomass gain” reported only density-normalized (kg/m3), not as absolute per-fish weight (g) or absolute biomass (kg) as the schema columns require.
- Total Feed referenced only symbolically inside the FCR formula, no absolute kg value stated.
- Daily weight gain (g/day) — a rate, not the cumulative per-fish gain the “Fish weight gain” column expects.
- Feed proximate composition beyond crude protein: 6.5% ether extract, 4% crude fiber, 14% mineral matter, 12% moisture.
- Water renewal rate stated as “three discharges/renewals per hour” (5 min drain, 14 min fill) — no L/min flow rate given.
- Per-species vegetable performance (lettuce/watercress/arugula leaves, roots, biomass) — trials.csv provides only one set of plant-performance columns per row for what is here three co-grown species; per-species breakdown recorded in remarks, with Total vegetable biomass (kg/m2) used to populate the “AP” column as the closest single yield metric.
Type classification rationale: experiment — randomized-block design with defined treatments (fish-tank volume ratio), true replication (3 blocks), and formal statistical testing (ANOVA/GLM, Student’s T test). No review or modelling component.
New tags introduced: Meta/Plant/Watercress, Meta/Plant/Arugula (new — no existing facet for either in the vault; checked against Meta/Plant/Lettuce, Meta/Plant/Chicory, Meta/Plant/Rocket, Meta/Plant/Basil, Meta/Plant/Cucumber, Meta/Plant/Tomato, Meta/Plant/Parsley, Meta/Plant/Barley, Meta/Plant/Coriander, Meta/Plant/Indian). Meta/Type/Experiment, Meta/Region/South-America, Meta/Fish/Tilapia all reused from existing vault entries (e.g. lenzCommonChicoryProduction2021, silvaAquiculturaManejoAproveitamento2013).
Quality tally: 0 ⚠️BLOCK, 1 ⚠️MATERIAL (Fish trial duration), 2 ⚠️CHECK (NO2-N basis; whole-vegetable weight attribution — CHECK does not count toward quality score). Score: ok (0 BLOCK and ≤2 MATERIAL).
Source: Barbosa et al. - 2020 - Performance of Nile Tilapia and vegetables Grown in Different Aquaponic Volumes.pdf
Data Tables
Structured data extracted from this paper into the vault's
trials.csv/plant_measurements.csvdatasets. Fields the paper didn't report are omitted. Download the full datasets (measurements).
Trial Parameters
barbosaPerformanceNileTilapia2020-T1
Fish
| Field | Value |
|---|---|
| Fish | Nile tilapia (Oreochromis niloticus) |
| Initial Stock density | 4 kg/m3 (12 fish/m3) (p.500) |
| FCR | 1.68 |
| Protein | 32 |
| Feed routine | Fed to apparent satiety, twice daily (08h00 and 17h00) (p.500) |
| Feed regime | Commercial extruded feed, 6-8 mm pellet (32% crude protein, 6.5% ether extract, 4% crude fiber, 14% mineral matter, 12% moisture) (p.500) |
| Fish survival rate | 95.64 |
| Fish trial duration (days) | ⚠ 40 |
Water
| Field | Value |
|---|---|
| Water volume in the system | 500 (fish tank only; total system incl. 500-L vegetable tank = 1000 L) (p.499) |
| Aq pH | 7.50 |
| Dissolved Oxigen | 4.37 |
| Water temperature | 24.96 |
| TAN / NH4-N | 0.42 |
| NO2-N | ⚠ 0.05 |
Plant
| Field | Value |
|---|---|
| Plant | Lettuce (Lactuca sativa), watercress (Rorippa nasturtium-aquaticum), arugula (Eruca sativa) |
| Details | Three vegetable species co-grown per unit: 20 lettuce + 40 watercress + 40 arugula seedlings in a shared 500-L gravel bed (2 m2 surface) (p.500) |
| Plants/m2 | 50.0 |
System & Setup
| Field | Value |
|---|---|
| System type | Media-filled bed (gravel), flood-and-drain via Bell siphon (p.499) |
| Media Details | Gravel size 1.5-2 cm, bed height 25-27 cm, 500-L tank with 2-m2 surface (p.499) |
| Biological system already in use | Y (Water previously matured with bacteria and algae was inoculated to stabilize the aquaponic systems (p.500)) |
| Air supplement | Y (Continuous aeration to each Nile tilapia tank via electromagnet air compressor (160 L/h) (p.499-500)) |
| Nutrient supplemented | Y (Fertilizer urea used to stabilize the aquaponic systems (p.500); [unclear] whether applied only pre-trial during cycling or continued during the 40-day data-collection period) |
| Equipment | 44-W submerged pump (JAD FP-58); Bell siphon (32-mm PVC riser, 50-mm bell, 100-mm media-guard); electromagnet air compressor (160 L/h); YSI Professional probes (temperature/DO/pH); Alfakit colorimetric kits (ammonia/TAN/nitrite) (p.499-500) |
| Control Parameters | Water temperature, DO and pH measured daily; ammonia (NH3), TAN and nitrite measured twice weekly; solids from feed/feces siphoned; evaporation losses replaced weekly (p.500) |
| Combination | Nile tilapia co-grown with lettuce, watercress and arugula in a shared gravel media bed; two aquaponic-only treatments compared by fish-tank volume (500 L vs 1000 L) at a fixed 500-L vegetable growing bed and fixed fish density (4 kg/m3, 12 fish/m3); no hydroponic-only control arm; randomized-block design, 3 blocks (periods) |
Site
| Field | Value |
|---|---|
| Region | South-America |
| Country | Brazil |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | kg/m3 (fish performance); kg/m2 (vegetable biomass, per species and total); g (lettuce/arugula leaves and roots, individual) |
| Statistic Details | ANOVA with treatment and block as sources of variation and fish initial weight as co-variable; GLM procedure, SAS 9.0; Student’s T test, P>0.05 = not significant (p.500-502) |
| Statistically analysed | Y |
| Replicates (n) | 3 |
| AP | 3.44 |
Experimental Remarks: TRIAL DEFINITION: T1 = fish tank-to-vegetable bed volume ratio 1:1 (500-L fish tank : 500-L vegetable/gravel tank), aquaponic-only treatment (no hydroponic control). Paired against T2 (1:2 ratio, 1000-L fish tank) in the same randomized-block design (3 blocks/periods); fish stocking density held constant at 4 kg/m3 (12 fish/m3) and vegetable bed volume/planting density held constant (500 L, 50 seedlings/m2) across both trials — only fish-tank volume differs (p.499-500). | WARN-MATERIAL Fish trial duration: Methods p.500 names the biometric-measurement window as ‘(40 days) of the experimental period,’ but the same paragraph’s daily-weight-gain formula divides by ‘30 days’: ’[(Final weight, g - Initial weight, in g)/30 days]’. Every other reference to trial length in the paper (Methods framing, Abstract, Conclusion) is consistent with 40 days; the 30-day figure appears only inside this one formula. Recorded 40 days as Fish trial duration; the 30-day divisor is very likely a drafting error, but cannot be confirmed, so it is flagged rather than silently corrected. Table 2’s Daily weight gain values (2.55 g/day T1, 2.77 g/day T2) are recorded as directly stated, not recomputed. Affects: interpretation of the daily-weight-gain rate basis only; does not affect other Table 2 cells. | WARN-CHECK NO2-N: Table 1 (p.501) labels the variable ‘Nitrite (mg L-1)’ using the ion notation ‘NO2-’, not explicitly ‘NO2-N’ (nitrite-nitrogen). The paper never states whether the reported mg/L basis is the NO2- ion or its N-equivalent (~3.28x apart). Recorded as stated (0.05 T1 / 0.09 T2) with no conversion applied, since the paper gives no basis to convert. Flagged for user verification. | WARN-CHECK Plant fresh weight (whole-vegetable weight): Discussion (p.503) states lettuce whole-vegetable weight ‘(80 and 121 g)’ when comparing to Sace and Fitzsimmons (2013), and a whole-arugula weight of ‘14 g’ comparing to Lennard and Ward (2019). Neither passage labels which figure belongs to T1 (1:1) vs T2 (1:2), nor whether 14 g is a pooled/overall mean across both treatments. The paper’s tables consistently list the 1:1 column before the 1:2 column, which would suggest 80 g = T1 and 121 g = T2, but this is an inference from formatting convention, not a stated label. Both trials’ ‘Plant fresh weight’ cells are recorded NR rather than guessed; raw values preserved here for verification. | NO COLUMN Ammonia (NH3, un-ionized): 0.02 mg/L, both T1 and T2, CV 18.84%, P=0.42 (Table 1, p.501) — no schema column distinct from TAN/NH4-N. | NO COLUMN Coefficients of variation (CV%) reported for every Table 1-3 variable are not converted to SD (computing SD from a reported CV would be derivation); trial means only are recorded in the numeric columns. CV%s: water temp 0.33, pH 0.19, DO 10.82, TAN 19.58, NH3 18.84, nitrite 21.54 (Table 1); fish final weight 2.05, FCR 16.67, survival 0.66, daily weight gain 11.31, final biomass 1.43, biomass gain 17.27 (Table 2); lettuce leaves 29.55, lettuce roots 17.50, lettuce biomass 18.87, watercress biomass 15.00, arugula leaves 27.42, arugula roots 26.12, arugula biomass 86.87, total biomass 26.96 (Table 3). | NO COLUMN Fish performance reported density-normalized (kg/m3) rather than as absolute per-fish weight (g) or absolute biomass (kg) as the schema’s ‘Fish size final’ and ‘Fish biomass created’ columns require; those columns are NR. Table 2 (p.501) values — T1: Final weight 0.436 kg/m3, Final biomass 4.71 kg/m3, Biomass gain 0.88 kg/m3; T2: Final weight 0.443 kg/m3, Final biomass 4.60 kg/m3, Biomass gain 0.71 kg/m3. | NO COLUMN Total Feed not reported as an absolute kg amount; it appears only as a symbolic term inside the apparent-feed-conversion formula (p.500). | NO COLUMN Daily weight gain (g/day), Table 2: T1 2.55, T2 2.77 — this is a rate, not the cumulative per-fish gain the ‘Fish weight gain’ schema column expects, so it is not entered there. | NO COLUMN Feed proximate composition (declared, same feed both trials): 32% crude protein, 6.5% ether extract, 4% crude fiber, 14% mineral matter, 12% moisture (p.500); only crude protein has a schema column (Protein). | NO COLUMN Water renewal rate: ‘three water discharges per hour’ / vegetable tank drains in 5 min and refills in 14 min (p.499-500) — no L/min flow rate stated, so Water recycle is NR. | NO COLUMN Per-species vegetable performance (Table 3, p.502, all P>0.05): Lettuce leaves (g) T1 64 / T2 90; lettuce roots (g) T1 16 / T2 31; lettuce biomass (kg/m2) T1 1.62 / T2 1.42; watercress biomass (kg/m2) T1 1.30 / T2 1.80; arugula leaves (g) T1 11 / T2 20; arugula roots (g) T1 4 / T2 6; arugula biomass (kg/m2) T1 0.51 / T2 0.64; total vegetable biomass (kg/m2) T1 3.44 / T2 4.88. Trials.csv provides only one set of plant-performance columns per row for three co-grown species, so these per-species values are recorded here rather than forced into Plant height/Leaf count/Plant fresh weight. The ‘AP’ column is populated with the Total vegetable biomass (kg/m2) figure as the closest available single yield metric for this row. | NO COLUMN Site coordinates for Campo Grande, MS are not stated anywhere in the paper; Lat/Long left NR rather than filled from outside knowledge of the city’s coordinates, per the prime directive.
barbosaPerformanceNileTilapia2020-T2
Fish
| Field | Value |
|---|---|
| Fish | Nile tilapia (Oreochromis niloticus) |
| Initial Stock density | 4 kg/m3 (12 fish/m3) (p.500) |
| FCR | 1.50 |
| Protein | 32 |
| Feed routine | Fed to apparent satiety, twice daily (08h00 and 17h00) (p.500) |
| Feed regime | Commercial extruded feed, 6-8 mm pellet (32% crude protein, 6.5% ether extract, 4% crude fiber, 14% mineral matter, 12% moisture) (p.500) |
| Fish survival rate | 96.01 |
| Fish trial duration (days) | ⚠ 40 |
Water
| Field | Value |
|---|---|
| Water volume in the system | 1000 (fish tank only; total system incl. 500-L vegetable tank = 1500 L) (p.499) |
| Aq pH | 7.52 |
| Dissolved Oxigen | 4.07 |
| Water temperature | 25.12 |
| TAN / NH4-N | 0.45 |
| NO2-N | ⚠ 0.09 |
Plant
| Field | Value |
|---|---|
| Plant | Lettuce (Lactuca sativa), watercress (Rorippa nasturtium-aquaticum), arugula (Eruca sativa) |
| Details | Three vegetable species co-grown per unit: 20 lettuce + 40 watercress + 40 arugula seedlings in a shared 500-L gravel bed (2 m2 surface) (p.500) |
| Plants/m2 | 50.0 |
System & Setup
| Field | Value |
|---|---|
| System type | Media-filled bed (gravel), flood-and-drain via Bell siphon (p.499) |
| Media Details | Gravel size 1.5-2 cm, bed height 25-27 cm, 500-L tank with 2-m2 surface (p.499) |
| Biological system already in use | Y (Water previously matured with bacteria and algae was inoculated to stabilize the aquaponic systems (p.500)) |
| Air supplement | Y (Continuous aeration to each Nile tilapia tank via electromagnet air compressor (160 L/h) (p.499-500)) |
| Nutrient supplemented | Y (Fertilizer urea used to stabilize the aquaponic systems (p.500); [unclear] whether applied only pre-trial during cycling or continued during the 40-day data-collection period) |
| Equipment | 44-W submerged pump (JAD FP-58); Bell siphon (32-mm PVC riser, 50-mm bell, 100-mm media-guard); electromagnet air compressor (160 L/h); YSI Professional probes (temperature/DO/pH); Alfakit colorimetric kits (ammonia/TAN/nitrite) (p.499-500) |
| Control Parameters | Water temperature, DO and pH measured daily; ammonia (NH3), TAN and nitrite measured twice weekly; solids from feed/feces siphoned; evaporation losses replaced weekly (p.500) |
| Combination | Nile tilapia co-grown with lettuce, watercress and arugula in a shared gravel media bed; two aquaponic-only treatments compared by fish-tank volume (500 L vs 1000 L) at a fixed 500-L vegetable growing bed and fixed fish density (4 kg/m3, 12 fish/m3); no hydroponic-only control arm; randomized-block design, 3 blocks (periods) |
Site
| Field | Value |
|---|---|
| Region | South-America |
| Country | Brazil |
Results & Statistics
| Field | Value |
|---|---|
| Measured Unit | kg/m3 (fish performance); kg/m2 (vegetable biomass, per species and total); g (lettuce/arugula leaves and roots, individual) |
| Statistic Details | ANOVA with treatment and block as sources of variation and fish initial weight as co-variable; GLM procedure, SAS 9.0; Student’s T test, P>0.05 = not significant (p.500-502) |
| Statistically analysed | Y |
| Replicates (n) | 3 |
| AP | 4.88 |
Experimental Remarks: TRIAL DEFINITION: T2 = fish tank-to-vegetable bed volume ratio 1:2 (1000-L fish tank : 500-L vegetable/gravel tank), aquaponic-only treatment (no hydroponic control). Paired against T1 (1:1 ratio, 500-L fish tank) in the same randomized-block design (3 blocks/periods); fish stocking density held constant at 4 kg/m3 (12 fish/m3) and vegetable bed volume/planting density held constant (500 L, 50 seedlings/m2) across both trials — only fish-tank volume differs (p.499-500). | WARN-MATERIAL Fish trial duration: Methods p.500 names the biometric-measurement window as ‘(40 days) of the experimental period,’ but the same paragraph’s daily-weight-gain formula divides by ‘30 days’: ’[(Final weight, g - Initial weight, in g)/30 days]’. Every other reference to trial length in the paper (Methods framing, Abstract, Conclusion) is consistent with 40 days; the 30-day figure appears only inside this one formula. Recorded 40 days as Fish trial duration; the 30-day divisor is very likely a drafting error, but cannot be confirmed, so it is flagged rather than silently corrected. Table 2’s Daily weight gain values (2.55 g/day T1, 2.77 g/day T2) are recorded as directly stated, not recomputed. Affects: interpretation of the daily-weight-gain rate basis only; does not affect other Table 2 cells. | WARN-CHECK NO2-N: Table 1 (p.501) labels the variable ‘Nitrite (mg L-1)’ using the ion notation ‘NO2-’, not explicitly ‘NO2-N’ (nitrite-nitrogen). The paper never states whether the reported mg/L basis is the NO2- ion or its N-equivalent (~3.28x apart). Recorded as stated (0.05 T1 / 0.09 T2) with no conversion applied, since the paper gives no basis to convert. Flagged for user verification. | WARN-CHECK Plant fresh weight (whole-vegetable weight): Discussion (p.503) states lettuce whole-vegetable weight ‘(80 and 121 g)’ when comparing to Sace and Fitzsimmons (2013), and a whole-arugula weight of ‘14 g’ comparing to Lennard and Ward (2019). Neither passage labels which figure belongs to T1 (1:1) vs T2 (1:2), nor whether 14 g is a pooled/overall mean across both treatments. The paper’s tables consistently list the 1:1 column before the 1:2 column, which would suggest 80 g = T1 and 121 g = T2, but this is an inference from formatting convention, not a stated label. Both trials’ ‘Plant fresh weight’ cells are recorded NR rather than guessed; raw values preserved here for verification. | NO COLUMN Ammonia (NH3, un-ionized): 0.02 mg/L, both T1 and T2, CV 18.84%, P=0.42 (Table 1, p.501) — no schema column distinct from TAN/NH4-N. | NO COLUMN Coefficients of variation (CV%) reported for every Table 1-3 variable are not converted to SD (computing SD from a reported CV would be derivation); trial means only are recorded in the numeric columns. CV%s: water temp 0.33, pH 0.19, DO 10.82, TAN 19.58, NH3 18.84, nitrite 21.54 (Table 1); fish final weight 2.05, FCR 16.67, survival 0.66, daily weight gain 11.31, final biomass 1.43, biomass gain 17.27 (Table 2); lettuce leaves 29.55, lettuce roots 17.50, lettuce biomass 18.87, watercress biomass 15.00, arugula leaves 27.42, arugula roots 26.12, arugula biomass 86.87, total biomass 26.96 (Table 3). | NO COLUMN Fish performance reported density-normalized (kg/m3) rather than as absolute per-fish weight (g) or absolute biomass (kg) as the schema’s ‘Fish size final’ and ‘Fish biomass created’ columns require; those columns are NR. Table 2 (p.501) values — T1: Final weight 0.436 kg/m3, Final biomass 4.71 kg/m3, Biomass gain 0.88 kg/m3; T2: Final weight 0.443 kg/m3, Final biomass 4.60 kg/m3, Biomass gain 0.71 kg/m3. | NO COLUMN Total Feed not reported as an absolute kg amount; it appears only as a symbolic term inside the apparent-feed-conversion formula (p.500). | NO COLUMN Daily weight gain (g/day), Table 2: T1 2.55, T2 2.77 — this is a rate, not the cumulative per-fish gain the ‘Fish weight gain’ schema column expects, so it is not entered there. | NO COLUMN Feed proximate composition (declared, same feed both trials): 32% crude protein, 6.5% ether extract, 4% crude fiber, 14% mineral matter, 12% moisture (p.500); only crude protein has a schema column (Protein). | NO COLUMN Water renewal rate: ‘three water discharges per hour’ / vegetable tank drains in 5 min and refills in 14 min (p.499-500) — no L/min flow rate stated, so Water recycle is NR. | NO COLUMN Per-species vegetable performance (Table 3, p.502, all P>0.05): Lettuce leaves (g) T1 64 / T2 90; lettuce roots (g) T1 16 / T2 31; lettuce biomass (kg/m2) T1 1.62 / T2 1.42; watercress biomass (kg/m2) T1 1.30 / T2 1.80; arugula leaves (g) T1 11 / T2 20; arugula roots (g) T1 4 / T2 6; arugula biomass (kg/m2) T1 0.51 / T2 0.64; total vegetable biomass (kg/m2) T1 3.44 / T2 4.88. Trials.csv provides only one set of plant-performance columns per row for three co-grown species, so these per-species values are recorded here rather than forced into Plant height/Leaf count/Plant fresh weight. The ‘AP’ column is populated with the Total vegetable biomass (kg/m2) figure as the closest available single yield metric for this row. | NO COLUMN Site coordinates for Campo Grande, MS are not stated anywhere in the paper; Lat/Long left NR rather than filled from outside knowledge of the city’s coordinates, per the prime directive.